The p:group element acts as a container for other instructions. It allows to distinguish between e.g. different queries or navigation tree results if you would have more than one of them.
<p:group id="expr" catchErrors="true|false"> [... child instructions ...] </p:group>
The id attribute is required.
The catchErrors attribute is optional. When this attribute has the value true, any errors that occur during the processing of the children of the p:group element will be caught. The result will then be a p:group element with an attribute error="true" and as child the stacktrace of the error. When catchErrors is true, the result of the execution of the children of the p:group element will need to be buffered temporarily, so only use this when you really need it.
A p:group element with id attribute, and contained in it the output of the child instructions.